IteratorDecorator 0.11

Adds dynamically interface of iterator to class

IteratorDecorator

Implementing iterator in Python is nothing complicated though what’s missing is possibility to make it even easier. This small library adds iter_attribute decorator allowing to quickly choose iterable for which iterator would be implemented.

Requirements

Python3.5+

Example

from IteratorDecorator import iter_attribute


@iter_attribute('number')
class CaseClass:
    def __init__(self):
        self.number = [1, 2, 3, 4]
        self.attr = ['attr1', 'attr2', 'attr3']

obj = CaseClass()

for num in obj:
    print(num)

Installing

In your virtualenv just call:

$ pip install IteratorDecorator

Warning

When using PyCharm or MYPY you’ll probably see issues with decorated class not being recognized as Iterator. That’s an issue which I could not overcome yet, it’s probably due to the fact that interpretation of object is being done statically rather than dynamically. MYPY checks for definition of methods in class code which changes at runtime. Since __iter__ and __next__ are added dynamically MYPY cannot find those defined in objects before object of a class is created. Possible workarounds for this issue are:

1. Define __iter__ method in class:

@iter_attribute('attr')
class Test:
    def __init__(self) -> None:
        self.attr = [1, 2, 3]

    def __iter__(self) -> 'Test':
        return self

Actually it does not have to be “real” __iter__ since it’ll be replaced by decorator implementation, but the definition is only needed for static checkers.

2. After creating object use cast or assert function denoting that particular instance inherits

from collections.Iterator:

assert isinstance(my_object, collections.Iterator)